Skip to content

Reason-Giving Template

A structured reasons artifact — instantiates Transparency for Accountability

A required structure that makes every decision state its grounds — the criterion applied, the material facts, and the conclusion — in a form a reviewer can check.

A Reason-Giving Template is a required structure that forces every decision to state its grounds in a checkable form — at minimum the criterion or rule applied, the material facts it was applied to, the conclusion, and (where a case was contested) why the losing considerations lost. Its distinguishing job among these mechanisms is to capture the rationale — the "why" — not the outcome, not the evidence file, not the rulebook. By making the reasoning a mandatory, structured field rather than an optional narrative, it converts "the committee decided" into "the committee decided because," in a form a reviewer can test point by point. It discloses reasons; it does not run the hearing — that whole adjudicative process belongs to a neighbouring pattern, Procedural Fairness.

Example

A parole board reviewing an inmate's release used to issue one-word outcomes — "denied." Under a reason-giving template, each decision must complete the same fields: the statutory criteria considered (risk to the public, conduct record, release plan), the specific facts under each (a clean disciplinary record for four years; an unverified housing plan), the decision, and the reason the negatives outweighed the positives this cycle. The applicant — and any reviewer — can now see not just that parole was denied but on what grounds, and can contest a ground that rests on a factual error. The duty to give reasons is what makes the denial challengeable.[1]

How it works

The distinguishing structure is a fixed reasoning skeleton applied to every decision, with required fields that cannot be left blank. The template ties the conclusion to a named criterion — which forces the decider to say which rule they used — and to the material facts — which exposes reasoning that rests on nothing. Its discipline is completeness of rationale, not readability or fairness of process: a terse, technical reason that is complete beats a fluent one that begs the question.

Tuning parameters

  • Required fields — what every decision must state (criterion, facts, conclusion, rejected alternatives). More fields force rigor but raise burden and boilerplate risk.
  • Granularity of grounds — one summary reason vs. a reason per issue. Finer supports precise challenge; coarser is faster.
  • Criterion linkage — whether each reason must cite the specific rule from the register. Tight linkage prevents invented standards; loose linkage is quicker.
  • Contested-issue handling — whether the losing arguments must be addressed. Addressing them is what makes the reasons honest; skipping them hides the close calls.
  • Standardization vs. fit — one template for all vs. per-decision-type variants. Standard forms compare well; variants fit unusual cases better.

When it helps, and when it misleads

Its strength is that a stated reason is a checkable reason: it exposes decisions that rest on a factual error or an unstated criterion, and it disciplines the decider in the very act of writing.

Its failure mode is that reasons can be manufactured after the fact — the template filled in to dress up a conclusion reached on other grounds, so a complete-looking rationale certifies a decision the reasons didn't actually drive. This is the "reasons as post-hoc justification" failure, the reasoning cousin of running a model backwards to ratify a choice already made. The discipline is to require the reasons at the moment of decision, tied to the specific criteria and facts on the record, and to let the challenge channel test them against the evidence — reasons that can't survive that test are the ones worth catching.

How it implements the components

  • decision_rationale_record — the mechanism's whole output: the structured record of the grounds for the decision.
  • rule_and_criteria_disclosure — each reason must name the specific criterion applied, tying the decision to a stated standard rather than an invented one.

It does not maintain the full catalogue of rules and standards (the Rule & Criteria Register; the template only cites the criterion in play), compile the underlying evidence (Evidence Disclosure Packet), or index the decision itself (the Public Decision Log).

Notes

Giving reasons is not the same as deciding fairly. A flawless rationale can document a biased or foreordained decision; the template makes the reasoning visible and challengeable, which is transparency's contribution — whether the process that produced it was fair belongs to Procedural Fairness, and whether the reasons are true is what the challenge channel and the evidence packet are for.

References

[1] The "duty to give reasons" is a well-established principle in administrative law: a body exercising power over someone should state the grounds for its decision, so the decision can be understood and challenged. It is the doctrinal anchor for this mechanism — reasons exist to be tested, not merely recorded. ↩